The pernicious effects of digital manipulation campaigns can reverberate through entire societies, but effectively detecting and responding to them is a challenge. Testing solutions in real-world contexts is highly complex and poses ethical barriers. Taking advantage of state-of-the-art approaches to modelling human social systems with face validity, in this project we overcome these barriers using a simulation approach. We have developed a scalable digital social environment simulator that offers fine-grained experimental control through precise and versatile configuration and a suite of evaluation analyses. We are using it to study information integrity in the context of societal events where it plays a central role, such as elections. By simulating real world and future manipulation strategies and analyzing their properties, we are pursuing a quantitative approach to prototyping effective defenses against them.
Selected Publications
A Simulation System Towards Solving Societal-Scale Manipulation
In this paper, we present a multi-agent simulator based on Deepmind’s Concordia, a software library for LLM-based multi-agent simulations of real world human experience. Our unique contribution is the addition of an online social media environment component as well as agent persona generation from measured psychological trait survey data. We demonstrate the simulator by simulating a 100-agent town election under different manipulation conditions. Through longitudinal surveys, we show via our custom dashboard the effects of manipulation in altering the outcome of elections.
Technology Transfer
Social Simulation
Multi-agent simulator based on Deepmind’s Concordia, a software library for LLM-based multi-agent simulations of real world human experience.